Articles and Posts

As a research project, I’m using generative AI to help me convert a production app built in elixir to node with typescript. I have no experience with functional programming, elixir, or typescript and haven’t touched node in six years. In 7 hours, I researched the new stack, got my project setup with: build script, testing, linting, coverage and complexity metrics. I’ve also setup the database by combining and converting the migrations scripts. Next I start migrating application code.

As LLM model providers have an arms race, we as consumers, can thread other considerations into our selection like transparency, training data, fair use, revenue model, energy consumption, privacy protection, security practices, open source, governance, and equitable access.

“The death of agile” will mean that average capabilities and values alignment of people still embracing it will go up. And the number of people using it as an excuse for their organizational dysfunction will go down. Hype is not necessarily true nor false. Neither is bullshit.

Unsurprisingly, if this study is to be believed. Return to office policies are highly correlated to male CEO’s whose income is disproportionately large compared to their employees and an increase in worker dissatisfaction. Not supported is the idea it does anything to improve company performance.

RTO doesn’t improve company value, but does make employees miserable: Study

For those investing and using generative AI as a tool for business. Navigate past the hype.

1. Own the material risks of such tools: personal / corporate privacy, data transparency / veracity, energy / compute resource consumption, regulatory compliance, and human safety.

2. Retain engineering and product capability to build with system reliability, product efficacy, business outcomes and great customer experience.

3. Lead humanely leveraging profit and operating efficiency towards better worker engagement / quality of life.

These are human accountabilities. We have a right to demand them of each other.

To develop my LLM based applications more responsibly, I’m setting them up with the option to run off local LLM and embedding models. That way I can compare results with different pre trained models and run pipelines in development without using cloud compute resources. It’s a tiny thing, I get it. But doing a small thing is better than doing no thing. #sustainability #responsibleTech

Leadership is taking risks to manage risk. You can’t know the pressure of that until you feel it. Regardless of our entirely human and ineffable mix of glories and failings, for that share of leaders who take that responsibility for other people’s well-being seriously, thank you.

I’ve built a tool backed by RAG to help support teams improve their operating runbook. It’s a combination chat agent and document database where problem solving completions based on manuals, threads, and tickets can be edited by an SME and saved as new run book entries which are then available to the team and improve the RAG responses.

Retrieval Augmented Generation can be a valuable and responsible use of generative AI.

RAG provides an agent built upon private or inaccessible content. It unhides information in ways that helps people get their jobs done.

Why RAG can be a responsible use of AI:

1. Immediately useful. It not only helps people access information, it can be part of a system to help them update and improve it. It can be introduced as a resource to existing workflows driving only as much change as an organization and its people can handle.

2. Lends itself to a human in the loop processes. It’s better at helping people do their jobs than automating their jobs.

3. Validations. As it’s built upon search, it can reference the sources used to generate the answer. This gives people a way to corroborate the information.

4. Privacy concerns. It can be entirely hosted within a secure perimeter.

5. Computing resources. RAGs leverage existing large language models without new large energy investments like training or fine tuning.

Spent a couple hours pairing with an end user of a software I built. We paired in the application doing what it was intended to do. Unfortunately, a rare experience.

We spend so much time disinter-mediated from the people who use the things we build. Whole organizations are established to preserve that distance. We dehumanize and alienate in the interest of efficiency. What we lose is a maker creating at tool to help a person accomplish a thing and for both to find joy in the act.

Rather than asking, “how can I use AI to replace a job.” Ask people who are very good at that job, “What part of your activities are necessary for you, your coworkers, or organization but takes time away from what you are uniquely here to accomplish?” Solve for that.

If this study’s findings are valid, for the time being, AI will supplement the high skill workforce, not displace it. At 25% it would basically right size supply and demand for tech talent.

That said, businesses will over react, reduce labor force more than they should and then rehire some as they better understand what AI can and can’t do. The caveat being, that tech is guilty of over hiring and overpaying in reaction to the growth opportunities of the pandemic.

So, who knows what actually ends up working to the job market. Because, we are not rational actors.